May 2000

DNA Computing

Continued from page 2

By Antonio Regalado

Nanotech C++

The enormous raw power of dna computing keeps the field moving in spite of all the daunting technical obstacles. Yet even if those obstacles ultimately prove insurmountable, Winfree's work could mean a breakthrough in the construction of ultrasmall devices. Indeed, Winfree himself thinks DNA tiles' most exciting application is as intelligent building blocks that put themselves together piece by piece on the nanometer scale-assembling into large and complex structures.

Collaborating with Rothemund and Adleman at USC, Winfree aims to fabricate a two-dimensional shape known as the Sierpinski triangle. Named after the Polish mathematician who discovered it in 1915, the triangle is a complex and beautiful fractal produced by repeating a simple geometric rule. The team plans to construct a real-world version of the triangle in a test tube using only seven different DNA tiles. Winfree has designed each tile type to carry out a simple program-to add itself to the growing shape or not, depending on the molecular cues provided by the triangle's outer edge.

In the hands of nanofabrication experts like NYU's Seeman, the DNA tiles could lead to easier methods to make exotic molecular structures-doing for nanotech what CAD and pre-fab building materials have done for the construction industry. "Greater control leads to things that you almost can't imagine," says Seeman. "Our expectation is that this approach can be applied to making designer materials and interesting patterns much more economically."

Seeman's lab, for instance, is already trying to attach nanoparticles of gold to DNA tiles in order to prototype tiny electrical circuits. These DNA "assemblies" would be about 10 times smaller than the tiniest features etched in silicon chips. However, Rothemund notes that there are limits to the patterns "computable" with DNA tiles. "We can't make anything we want," says Rothemund. "But the simple assemblies we've made so far show how well the basic operations work."

They also show how much scientists still have to learn. Winfree compares his efforts so far to one-line programs written in biochemical Basic. What he'd really like to be doing is programming biochemical reactions in C++. He expects this more advanced language will evolve as researchers master new operations, such as selectively removing tiles from an assembly. Winfree speculates that one day it may be possible to bring this growing repertoire of programmable components together to build synthetic systems-call them "nanorobots" if you wish-capable of independently carrying out useful tasks. "The really interesting direction DNA computing is taking us is to see just how far we can learn to program biochemical reactions," says Winfree.

That may sound like futuristic hype, but researchers are already beginning to figure out ways to do it. At Lucent Technologies' Bell Labs, physicist Bernie Yurke, for one, is working with DNA in the hopes of assembling ultrasmall molecular motors. Yurke imagines that some day it might be possible to build a DNA motor that could walk across Winfree's DNA tiling constructs, making chemical changes at specific points. "You could lay down an arbitrarily complex pattern," Yurke says, which might then be transferred to a silicon substrate to fabricate nanometer-scale circuits and transistors. "My hope is that in the future complicated electronic structures like computers will be made this way."

Electronic computers assembled using DNA that computes? It may sound like an unlikely twist in the evolution of DNA computing, but it's one that Adleman believes is entirely in keeping with the field he helped launch. "Like quantum computing, DNA computing is very futuristic, and both make the point that computation doesn't have to take place in the box that sits on our desks," says Adleman, this time in a telephone interview. "Even if they don't become a viable means of computing in the future-and I don't know if they will-we may learn what the real computer of the future should look like."

Computing (and Constructing) With DNA Organization Key Researchers Focus Bell Labs Bernie Yurke, Allan Mills Fabricating DNA motors for assembling electronic components Duke University/Caltech John Reif, Thomas LaBean, Erik Winfree (Caltech) Working on massively parallel addition using DNA tiles New York University Nadrian Seeman Assembling complex nanostructures out of DNA Princeton University Laura Landweber, Richard Lipton RNA-based computer used to solve chess puzzle known as the "knight problem" University of Southern California Leonard Adleman Automating a self-contained lab system for DNA computing; proved, in theory, that DNA can crack DES data encryption standard University of Wisconsin Robert M. Corn, Lloyd M. Smith, Anne E. Condon, Max G. Lagally Adapting DNA-chip technology to do DNA computation on a solid surface